The MIMS (Membrane Inlet or Interface Mass Spectrometry) method of sample introduction was first described by Hoch and Kok in 1963 and has been summarized in numerous recent articles. Most generally and according to this method, a semi-permeable membrane interface separates the gaseous or liquid sample matrix (e.g., carrier gas, ambient air, water) from the much lower pressure region that is created within a mass spectrometer. Analyte molecules pass through the semi-permeable membrane—the latter being typically formed solely from a silicone polymer—preferentially to the transport of molecules from a sample matrix. Thus and in comparison to mass spectrometry systems having direct sample inlets such as an orifice or capillary, a MIMS inlet can be operated such that it permits only a much smaller number of matrix molecules into the vacuum of the mass spectrometer for the same number of analyte molecules. This enables the construction of a mass spectrometry system having substantially the same sensitivity, but with a reduced gas load on the vacuum pumps responsible for maintaining the low pressure of the mass spectrometer. Resulting systems can be built with more robust vacuum pumps which require lower power, enabling the construction of smaller, more durable, and thus more portable mass spectrometer instruments.
MIMS instruments have demonstrated their largest commercial successes in the environmental and security fields. These applications typically require compact instruments featuring ruggedized and low-power vacuum pumping systems. Instruments of this type include, for example, the family of Mobile Mass Spectrometer instruments produced by Bruker Daltonik GmbH and the battery operated HAPSITE instruments developed by INFICON, Inc. as described in U.S. Pat. No. 5,426,300 to Voss et al. Both of these manufacturers use linear quadrupole mass filter detectors and optional gas chromatography sample introduction systems in order to provide a high sensitivity analysis for chemical warfare agents (CWA) and a wide range of toxic industrial chemicals (TICs) in a rugged package. Instruments such as those produced by Kore Technology Ltd also use a membrane inlet; however, these instruments make use of a different mass spectrometry—i.e., technology Time-of-Flight (TOF)—rather than a linear quadrupole mass filter. All of the above-noted systems take advantage of MIMS to enable the use of low speed sputter ion pumps or these pumps combined with small chemical getter pumps to enable portable or semi-portable operation. Numerous other implementations of MIMS instruments have been developed, although these have been largely relegated to research applications. One such system was developed for submersed water sampling, as described in U.S. Pat. No. 6,727,498 to Fries et al. In this latter device, the inlet membrane permits the introduction of analyte from a water matrix while withstanding the pressure of up to several hundred meters of water.
One of the major limitations of MIMS instruments stems from the reality that analyte molecules must diffuse thorough the semi-permeable membrane before entering the ion source of the mass spectrometer. Larger molecules such as pesticides, narcotics, some explosives, etc. diffuse though the membrane much more slowly than smaller molecules, such as many volatile organic chemicals (VOCs). Because of this limitation, MIMS systems—including the commercial systems mentioned above—have been largely relegated to the detection of VOCs or only the smallest semivolatile organic compounds (SVOCs). In general, SVOC compounds tend to be larger molecules than VOCs and thus diffuse more slowly through a MIMS inlet. Some success has been made in the detection of larger molecules by using much higher membrane temperatures in order to speed the diffusion process. One prior art approach involved heating the membrane directly with the photon flux from the ion source filament; however, there are limitations to using higher temperature as the sole approach. For many analytes, high temperature reduces sensitivity by lowering the analyte partition coefficient into the membrane, and for silicone membranes—the most prevalent type of semi-permeable membrane—contact with small amounts of water or oxygen at elevated temperature causes liberation of volatile silicon containing compounds. These liberated materials coat and degrade the electrode surfaces of the mass spectrometer. Additionally, higher temperature may cause labile analytes to degrade.
Moreover and even at elevated temperatures, MIMS systems typically have time responses that are too slow for use as a gas chromographic detector in applications requiring the analysis of SVOCs and other large molecules. Slow diffusion of large molecules causes peak broadening, which leads to both overlap and a reduced signal to noise ratio. The use of thinner membranes would enable faster diffusion without the problems associated with excessively high temperatures, and due to the squared dependence of diffusion time on membrane thickness, the use of thinner membranes would theoretically have a marked improvement on the time response of MIMS systems used for the analysis of SVOCs.
Although thin membranes are expected to enable higher performance MIMS in many instances, to find use in MIMS, these devices must overcome several impediments. First, the membrane must be capable of withstanding a high pressure drop without rupture. Second, the membrane must be free from pinholes or other similar defects. Third, the membrane must be constructed from materials that exhibit favorable sorption and diffusion behavior for the analyte of interest while also tolerating elevated temperatures, as required for many applications. Fourth, the membrane must not outgas vapors which themselves interfere with the detection of analyte or otherwise degrade the detection equipment.
Semi-permeable membranes are typically rubbery polymers with limited mechanical strength, yet it is essential that they withstand large pressure differentials between the matrix and the vacuum of the mass spectrometer. U.S. Pat. No. 5,703,359 to Wampler teaches a method of incorporating fibers into the membrane to improve its strength; however, it is often necessary to increase the mechanical stability of the semi-permeable membrane by use of a rigid support structure between the membrane and the vacuum of the mass spectrometer. To that end, a number of support structures have been used, wherein examples include wire mesh, photo-etched metal, deep reactive ion etched silicon, and sintered frits. In general, the smaller the size of the hole in the support structure, the better this structure supports the membrane. This technique enables the use of much thinner membranes. Improved support also enables operation with larger pressure differentials between the matrix and the mass spectrometer, which is desirable in many applications such as underwater sampling. Support structures not only prevent catastrophic rupture, but they also prevent deformation and resulting hysteresis behavior.
One candidate support material is anodized or anodic aluminum oxide (AAO). Briefly, sheets of AAO are formed by chemical etching of thin aluminum foils or films under high electric fields. A unique property is that wafers of this material can be prepared with a high density of extremely small channels (5-100 nm in diameter) extending across the support structure from one face to the other. The openings of these narrow diameter channels are sufficiently small that they enable the use of ultra-thin membranes. Using AAO, membranes could be far thinner than prior known membranes having support structures which have nominal pore sizes larger than 100 nm. The direct pathway from one face to the other also provides a more ideal flow path than the tortuous pathway through a sintered frit; however, sintered fits could in principle be composed from oriented fibers in order to produce oriented channels. Processes, such as deep reactive ion etching, can also create substrates with parallel channels, although these are of somewhat larger feature sizes. Another advantage of AAO is that this material is chemically inert and tolerant to high temperatures (>400° C.). AAO substrates are sometimes referred to as membranes; however, to avoid confusion with the semi-permeable polymeric films which actually serve the function of membranes herein, we herein generally refer to AAO material by its function herein, which is as a support structure in order to enable the use of thin semi-permeable membranes in MIMS.
AAO has found many uses, including some limited applications in MIMS. In one prior implementation, carbon nanotubes were deposited into the pores of the AAO material and used in a gas flow arrangement in which the pores are analogous to a large number of parallel capillaries. There was no thin semi-permeable membrane such as PDMS requiring solvation of analytes, followed by transport though the membrane and then release on the high vacuum side at the ion source of the mass spectrometer. A similar prior art approach in the development of a gas separation filter employs a first step of narrowing the pores of the AAO substrate and then coating the surface with a microporous gas separation layer of an inorganic polymer. In this latter case, the pores of AAO were filled with an inorganic material to narrow the pores and then spin coated over the top to create a microporous membrane. This top layer was not a semi-permeable layer such as that preferred for MIMS applications, and there was no mention of using this for the inlet of a mass spectrometer. In yet another prior art implementation, polymer films were coated onto AAO membranes, but no technique was provided to prevent this material from penetrating into the pores/channels as would be required to keep the membrane thin. The focus of this latter membrane was for use in a preparatory filtration of biological samples, such as low molecular weight proteins wherein use for MIMS was not discussed or otherwise suggested.
AAO is commercially available in 13 mm diameter disks, with a thickness of 60 microns which is much thinner than that typically possible with sintered frits; however, it is obvious that the invention described herein also extends to other nano or microporous support structures such as sintered frits, or structures created with top down lithographic techniques, such as deep reactive ion etching.
As previously mentioned above, thin membranes lack the mechanical strength required to support large pressure drops between the matrix environment and the vacuum of the mass spectrometer. This lack of strength can cause deformation, which may result in unacceptable changes in device characteristics. Thinner membranes can rupture, resulting in a loss of vacuum and damage to the mass spectrometer. Supporting structures with small feature sizes are also important in that they enable the use of thinner films and weaker membranes (i.e., polymer films that are not cross-linked).
Several methods are known for producing thin polymer films which may be suitable for use as semi-permeable membranes. Well known examples include casting, spin coating, dip coating, and certain physical and chemical vapor deposition processes. The goal of these processes is to produce thin, defect-free films. Creating the film directly on the support structure is often required for thin films because they lack the mechanical resiliency to withstand a mechanical transfer from production surface to support. Mechanical assembly techniques which involve forming the film elsewhere and then moving it to the porous support structure are typically limited to strong films with a thickness in excess of 10 μm. These steps may also create thin areas and other defects in thin membrane assemblies.
A previously developed MIMS technique did make use of an extremely thin membrane for the analysis of SVOCs. This approach used 0.5 micron thick PDMS films sputtered onto microporous polypropylene support fibers. However, in this implementation, a second counter-flowing carrier gas was used on the distal side of the membrane to capture analyte traveling through the membrane. Thus and according to this design, the membrane did not serve to separate the high pressure on the matrix side from the low pressure vacuum of the mss spectrometer. This system would thus fail to benefit from the largest advantage of a true MIMS system, which is a greatly reduced gas load on the pumping system compared to conventional inlet systems using capillaries or orifices. It is unlikely that the microporous support structure described by this latter approach would have the same strength as the nanoporous assemblies described below because the distance spanned by the membrane to cross a microscale gap is obviously larger than that to span a nanoscale gap. Additionally, it seems unlikely that polypropylene fiber would be able to withstand high temperature operation, and it would not be as chemically inert as desired for many MIMS applications.